Multi-organ involvement assessed by MRI in patients with idiopathic

Multi-organ Involvement Assessed by MRI in Patients with Thrombotic Thrombocytopenic
Giovanni Quarta1,2, Marie Scully3, Hyare Harpreet4, Andrew S. Flett1, Jim Gilbert5, William J
McKenna1, James C Moon1
1) The Heart Hospital, University College London Hospitals Trust, London, UK
2) Department of Cardiology, S. Andrea Hospital, University “La Sapienza”, Rome, Italy
3) Department of Haematology, University College London Hospitals Trust, London, UK
4) Imaging Department, University College London Hospitals Trust, London, UK
5) Archemix Corporation
Background: Thrombotic Thrombocytopenic Purpura (TTP) is an uncommon, acquired, life
threatening, autoimmune disorder characterised by multi-organ involvement with microvascular
occlusion. Although autopsy studies have shown high incidence of cardiac involvement, clinical
evidence of cardiac disease is limited.
Hypothesis: We hypothesised that, in TTP: 1) CMR would detect otherwise unrecognized cardiac
involvement and 2) this involvement would correlate with other organ involvement.
Method: Thirteen consecutive patients (4 males, 9 females, mean age: 47 ± 13 years), enrolled as part
of an observational study, were evaluated with standard cardiac and brain magnetic resonance. The late
gadolinium enhancement (LGE) technique was used to assess areas of cardiac focal fibrosis potentially
caused by micro-thrombotic damage.
Findings: All patients had normal LV ejection fraction (mean EF 66 ± 5%). LV volumes and mass
were normal in 11 of 13 patients and raised in 2. No patient had regional wall motion abnormalities.
Three patients (23%) had patches of LGE which were sub-endocardial and limited in size (Figure 1a
and 1b). By contrast, only two patients had a completely normal brain MRI. Four had supratentorial
white matter lesions; seven had cerebellar/deep grey matter/subcortical/cortical infarcts (Figure 1c and
1d). All three patients with LGE had severe brain involvement. The extent of brain damage correlated
with the presence of cardiac LGE (Kendall's tau-c = 0.43, p=0.023).
Conclusion: In this pilot study looking at cardiac involvement in TTP by CMR, CMR found LGE
patches in one quarter of patients (without wall motion abnormalities). LGE was more likely to be
found if there was more severe brain involvement.
Figure 1a and 1b. LGE images from a patient with TTP, showing a patchy LGE in the infero-lateral
wall and papillary muscle (arrows). The patient had a normal coronary angiogram. c) Axial FLAIR
sequence from an MRI brain study in the same patient demonstrates multiple cortical and subcortical
areas of ischaemic damage (arrows). d) Axial FLAIR image showing a mature left middle cerebral
artery branch ischaemic infarct involving the left parietal lobe, associated with mature blood products